Despite of having too much lighting this is not a gaming rig. I use this system primarily for machine learning as well as algorithm development. These kind of workloads guided choice of components, Optane SSDs for search algorithms and Titan RTXs for machine learning. I could not use 2080 Ti, not enough memory. Going with Quadro 6000 would not give much but significantly more expensive.
This system was originally running with air cooling. Some of machine learning tasks take many days to complete with GPUs running close to 100%. On air they would get very hot and were throttling. Throttling would add more time to already very long runs. This led me to upgrade to custom loop. I've done couple of custom loops before but they were on soft tubes. This is my first hard tube build and it is far from perfect from esthetics point of view. But it cuts running time for ML tasks by about 20 percent which is quite significant as I can do more runs and get better models.
My cable management skills are non existent so I will not show back of the case. But there are so many cables: power, fans, temperature and flow sensors, aquabus, usb, various leds, that i'm not sure how to handle them. I really like this case with one exception, there is only about an inch wide space behind the main compartment to fit all the cables in. I was not successful in achieving this, the case is open in the back for now.